Neuroendocrine tumors; measures to improve treatment and supportive care
de Hosson, Lotte Doortje
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
de Hosson, L. D. (2019). Neuroendocrine tumors; measures to improve treatment and supportive care.
Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
metastatic neuroendocrine tumors
based on [
18
F]FDOPA PET/CT
Lotte D. de Hosson
1, Aline M. van der Loo- van der Schaaf
2,
Ronald Boellaard
3, Johannes H. van Snick
3, Elisabeth G.E. de Vries
1,
Adrienne H. Brouwers
3, Annemiek M.E. Walenkamp
11Department of Medical Oncology, University of Groningen, University Medical Center
Groningen, Groningen, The Netherlands
2Department of Radiology, University Medical Center Groningen, University of
Groningen, Groningen. Current address, Medical Center Zuiderzee, Lelystad, The Netherlands
3Department of Nuclear Medicine and Molecular Imaging, University of Groningen,
Abstract
Purpose
Neuroendocrine tumors (NETs) can produce neuroendocrine amines resulting in symptoms. Selecting the most active amine-producing tumor lesions for local treat-ment might be beneficial for patients with metastatic small intestinal (SI-)NET. Tumor burden, correlates with catecholamine pathway activity. We analyzed interlesional heterogeneity with 6-[18F]fluoro-L-3,4-dihydroxyphenylalanine ([18F]FDOPA) Positron
Emission Tomography (PET) scans in patients with SI-NET and investigated if lesions with substantially higher [18F]FDOPA uptake could be identified.
Procedures
For all lesions on [18F]FDOPA PET scans the [18F]FDOPA uptake was calculated by
dividing standard uptake value (SUV) peak of the lesion by the SUV mean of the background organ. The magnitude of heterogeneity between lesions within a patient was calculated by dividing the lesion with highest by the one with lowest [18F]FDOPA
uptake. Lesions with a higher [18F]FDOPA uptake than the upper inner or outer fence
(>1.5 or 3 times the interquartile range above the third quartile) were defined as lesions with mild or extreme high [18F]FDOPA uptake, respectively and presence of these were
determined in patients with ≥10 lesions.
Results
With [18F]FDOPA, over 680 lesions in 38 patients, of which 35 serotonin producing,
were detected. [18F]FDOPA uptake varied with a median of 8-fold up to 44-fold between
lesions within a patient. In 12 of 20 evaluable patients lesions with mild high [18F]FDOPA
uptake were found and in five, lesions with extreme high [18F]FDOPA uptake.
Conclusions
[18F]FDOPA PET showed considerable heterogeneity in [18F]FDOPA uptake between
tu-mor lesions and identified lesions within patients with mild or extreme high [18F]FDOPA
uptake.
5
Introduction
Neuroendocrine tumors of small intestine (SI-NET) are rare tumors which in general are characterized by slow tumor growth. Over 50% of patients with NET present with metastatic disease. Symptoms are caused by tumor mass and/or (over-) production of neuroendocrine amines such as serotonin (1). Heterogeneity of tumor characteristics is increasingly appreciated and common in NETs and occurs at different genetic and epigenetic levels and in the expression of protein biomarkers (2). In several pathology studies heterogeneity between tumor lesions is shown. Immunohistochemical analysis of surgical specimens of liver metastases of 29 patients with pancreatic NET demon-strated heterogeneity within and between synchronous and metachronous metastases for Mindbomb E3 ubiquitin protein ligase 1, CD34, and somatostatin receptor 2 (SS-TRA2) (3). In another study, in 14 out of 27 patients with SI-NET a discordant tumor grade between liver lesions was demonstrated (4). This was also shown in pancreatic NET where 10 out of 16 patients had a discordant tumor grade between nodal, liver or other metastases (5). Furthermore eight out of 26 patients had tumor lesions with no to weak SSTRA2 expression and lesions with moderate to strong SSTRA2 expres-sion (6). For patients with SI-NET a number of systemic and local treatment options are available. Selection of the most active tumor lesion for local treatment might be beneficial for patients with SI-NET to reduce symptoms caused by overproduction of neuroendocrine amines. Local treatment options consists of radiofrequency ablation (RFA), transarterial chemo-embolization (TACE), transarterial embolization (TAE), resection, or the more recently developed, selective internal radiation therapy (SIRT) (7,8). In patients with colorectal liver metastases SIRT added to standard chemotherapy had a better median PFS in the liver than chemotherapy (9). Currently, no studies ana-lyzing heterogeneity between tumor lesions based on molecular imaging are available. In patients with SI-NET the tumor can be visualized with several molecular imaging techniques. [18F]FDOPA visualizes the metabolism of catecholamine pathway in the
tumor cell. Total tumor burden measured with [18F]FDOPA PET scans in patients with
SI-NET correlates with tumor markers of the serotonin and catecholamine pathway (10). We analyzed interlesional tumor [18F]FDOPA uptake heterogeneity with correction for
background activity with PET scans and investigated if tumor lesions with substantially higher [18F]FDOPA uptake could be identified in patients with SI-NET.
Materials and methods
Patients
Patient records of all patients with NET undergoing an [18F]FDOPA PET scan at the
University Medical Center of Groningen (UMCG), The Netherlands, between February 2014 until April 2015, were screened. All adult patients with a metastasized SI-NET with visualization of more than one lesion on the [18F]FDOPA PET scan and who had
under-gone a diagnostic computed tomographic (CT) scan within 6 months of the PET scan were selected for further analysis (Fig. S1). The following baseline characteristics were retrieved from patients medical record; duration of disease, previous treatments, tumor grade (1 or 2) according to the 2010 WHO classification, 5-HIAA level in 24-h urine and serotonin level in platelet rich plasma. A tumor was considered serotonin-producing if either a 5-HIAA level in 24-h urine >3.8 mmol/mol creatinine or a serotonin level >5.4 nmol/109 thrombocytes in platelet rich plasma was measured (11). [18F]FDOPA PET
scans were conducted during standard care. Because of the retrospective nature of this analysis, according to the Dutch regulations and the ethical committee of our institution, no approval by this committee was needed and no additional informed consent was required. The patient whose scan was used in Figure 2, gave informed consent for publication, according to author guidelines.
[
18F]FDOPA PET and CT Scan
[18F]FDOPA was produced in the radiochemical laboratory of the UMCG as described
previously (12). Patients fasted for 6 hours before the tracer injection and were allowed to continue all medication. For the reduction of tracer decarboxylation and subsequent renal clearance, all patients received 2 mg/kg carbidopa (maximum 150 mg) orally as pre-treatment, 1 hour before the [18F]FDOPA injection to increase [18F]FDOPA uptake in
tumor cells (13). PET images were acquired 60 min after intravenous administration of [18F]FDOPA (170-215 MBq). Scanning was performed from the upper legs to head with
a PET-CT camera (Siemens Biograph mCT 40 or 64 slices, 4 detector rings) with zoom factor 1. Scanning time per bed position was 1 to 3 min depending on body weight, and a low dose CT was used for attenuation and scatter correction. PET data were reconstructed with Siemens Ultra HD (trueX and time of flight), and ordered subsets expectation maximization reconstruction (OSEM), using 3 iterations and 21 subsets and a matrix of 400 with a full width at half maximum (FWHM) of 5 mm Gaussian (isotropic) filter (10,14,15). The PET scan was combined, at the same time or within 6 months before or after date of PET scanning, with a contrast enhanced diagnostic CT scan. A CT scan of the chest (n=32) and abdomen (n=38) was obtained, with iodine-containing intravenous, Iomeron 350 mg/mL, and oral contrast agents. CT scans were
5
performed on the Biograph mCT PET-CT, 40-slices, or 64-slices, Somatom Sensation 64, or Somatom Definition Dual (all Siemens) with a maximal slice thickness of 2 mm.
Imaging Data Analysis
PET-positive tumor lesions were defined as lesions with an unequivocal visibility on [18F]FDOPA PET scan greater than normal activity in that body region (16). Around all
PET-positive tumor lesions a volume of interest (VOI) was manually drawn on the PET scan with use of in-house developed Accurate Software. In each ‘background organ’, a spherical VOI was drawn in an area of homogenous activity in the organ, to quantify the physiological [18F]FDOPA uptake. The VOI diameter used to measure SUV mean
in background organs was 1 cm in blood pool, bone marrow at L4 or L5, cortical bone of the humerus, cortex of kidney and pancreas; the VOI diameter was 2 cm in muscle, myocardium and spleen; and the VOI diameter was 3 cm in brain, liver and lung. SUV max, SUV peak and SUV mean values were calculated according to the body weight of the patient.
To detect and quantify all tumor lesions, including PET-negative tumor lesions, [18F]FDOPA PET scans were compared with CT scans that were reviewed by a radiologist
with knowledge of the PET data. The radiologist counted in each organ all abnormalities and defined these as definitely benign, definitely malignant, or as ‘inconclusive lesions’. Afterwards, all discrepancies between malignant lesions or inconclusive lesions on CT scan versus tumor lesions on [18F]FDOPA PET scans were reviewed by one of the
investigators. If a PET-negative, definitely malignant or ‘inconclusive lesion’ was found, a VOI was drawn surrounding the lesion on CT scan to calculate SUV values of the same region on the [18F]FDOPA PET scans. For small PET-negative lesions <0.15 mL,
like lung nodules, a spherical VOI with a diameter of 1 cm was drawn. In case of more than 20 liver lesions on [18F]FDOPA PET or CT scan, a cut-off of 20 lesions for both
modalities was used. If > 20 liver lesions were present, single lesions could not be identified separately and therefore could not be counted exactly. PET-negative, CT defi-nitely malignant lesions were included for further analysis of [18F]FDOPA heterogeneity.
Lesions that were ‘inconclusive’ on CT and PET-negative were not used for further analysis of [18F]FDOPA heterogeneity.
Patients were evaluable for detection of lesions with mild or extreme high [18F]FDOPA
uptake if > 10 tumor lesions were present, see for calculation below. [18F]FDOPA uptake
in a tumor lesion was calculated by dividing the SUV peak value of the tumor lesion by the SUV mean value measured in the VOI drawn in the background organ. To determine the ‘tumor burden’ for each lesion or VOI the uptake has to be multiplied by lesion or VOI volume (10,17). This method was analyzed in a study using 18
fluorodeoxyglucose-PET scans. SUV peak corrected for local SUV mean background multiplied by volume had the most optimal feasibility and repeatability to measure metabolic active tumor
volume (i.e. tumor burden) (17). To explore the partial volume effect of the results, SUVs of tumor lesions were correlated with tumor VOI volume and the correlation was compared with the relation between SUV and volume seen in the NEMA NU 2-2007 phantom model (18). To explore effect of tissue density on SUV, the SUV of background organs was correlated with tissue density. Tissue density was based on the average Hounsfield Units per VOI.
Statistical Analysis
Patients’ demographic and clinical variables were summarized using medians with ranges for continuous variables and frequencies with percentages for categorical variables. The magnitude of heterogeneity between lesions within a patient was calcu-lated by dividing the lesion with highest [18F]FDOPA uptake with the lesion with lowest
[18F]FDOPA uptake within a patient. Differences between a median [18F]FDOPA uptake
or SUV mean per organ was analyzed with the Kruskal Wallis test. To determine lesions with mild or extreme higher [18F]FDOPA uptake than other lesions, the lower (Q1) and
upper quartiles (Q3) were determined first. The interquartile range (IQR), or Q3–Q1, was then computed. Lastly, the upper fences were computed as follows: upper inner fence Q3 +1.5* (IQR) and upper outer fence Q3 +3 * (IQR). Any data points outside of the fences, are considered as lesions with mild or extreme high [18F]FDOPA uptake,
compared to the other lesions in that patient. Correlations of parametric values were calculated using Pearson’s r, nonparametric values were calculated using Spearman’s rho. A p-value < 0.05 was considered significant. Statistical analyses were performed using SPSS version 23 (SPSS, Inc., Chicago, IL).
Results
Patients
In the selected period 182 [18F]FDOPA PET scans of 160 patients with SI-NET were
performed. Eighty three patients were excluded because their scan showed no or only one tumor lesion, 29 patients had a primary NET of pancreatic or lung origin and in 10 patients the time between [18F]FDOPA PET scan and CT scan was more than 6 months
(Fig. S1). In total 38 patients were eligible for analysis. Data of 5-HIAA urinalysis and serotonin in platelet rich plasma were available for 34 and 38 patients, respectively. According to these data, 35 patients had a serotonin producing NET. Baseline charac-teristics of included patients are shown in Table 1. Median age of the patients was 65 years (interquartile range 55-70), and 55% of patients were male.
5
Analysis of tumor lesions measured with [
18F]FDOPA PET
Over 680 tumor lesions were visualized on the [18F]FDOPA PET scans (mean > 20
lesions per patient). The median SUV peak in tumor lesions was 5.5 (interquartile range 3.1-11.2) and the median [18F]FDOPA uptake was 4.1 (interquartile range 2.2-7.7). In
total 581 [18F]FDOPA PET lesions could be correlated to a morphologic lesion on CT
scan (Fig. S2). Thirty eight [18F]FDOPA PET-negative lesions were detected by CT
scan. In one patient the CT scan revealed two liver metastases that were not detected on [18F]FDOPA PET scan and these were included for analysis of [18F]FDOPA
hetero-geneity. Of 36 other [18F]FDOPA PET-negative lesions detected by CT scan, mostly
localized in lung and lymph nodes, no differentiation could be made on CT whether the lesions were benign or malignant and these were not analyzed. In some patients one PET lesion referred to two or more CT lesions, because the lesions were merged with each other on [18F]FDOPA PET scan. Seven lesions were seen on the PET-scan but
outside the field of view on CT-scan.
Examination of the lesions per organ showed 91 bone lesions visible on [18F]FDOPA
PET scan, inside the view of field of the CT scan, but not visible on CT scan. When we evaluated other organs than the skeleton, a small number of lesions were revealed only on [18F]FDOPA PET scan (liver, lymph node/mesenterium, lung) or only on CT scan
(lung) (Fig. S2).
Interlesional heterogeneity and physiological [
18F]FDOPA uptake
The [18F]FDOPA uptake ranged from 1- to 44-fold between individual lesions within the
same patient (median 8-fold) (Table 2). Twelve of the 20 patients with over 10 tumor lesions had lesions with mild high [18F]FDOPA uptake and five patients had lesions
with extreme high [18F]FDOPA uptake (Fig. 1,2). Liver, lymph nodes combined with
Table 1. Baseline characteristics of small intestinal NET-patiens (n=38)
Characteristic n of patients (%) Sex male, 21 (55) Tumor grade: 1 2 Unknown 27 (71) 3 (8) 8 (21) Treatment: SSA use Surgery Everolimus Interferon PRRT Any treatment 23 (61) 22 (58) 1 (3) 3 (8) 1 (3) 26 (68) Patients with serotonin producing-tumor 35 (92)
n= number, NET=neuroendocrine tumors, PRRT=peptide radionuclide receptor therapy, SD=standard deviation, SSA=somatostatin analogue.
Table 2. Tracer uptake, magnitude of heterogeneity and number of lesions with mild or ex-treme high tracer uptake.
Characteristics of metastases Tracer uptake Pt Lesion (n) Metastasized organs Mean Max vs
min MH (n)* (n)EH * 1 >65 Lung, Liver, Pancreas, LN, Bone, Other 6.5 43.6 3 1
2 17 Liver 2.2 3.7 0 0 3 3 Liver, LN 1.5 1.2 - -4 38 Lung, Liver, LN 6.0 3.6 0 0 5 6 Bowel, LN 3.0 3.0 - -6 5 Lung, Liver 3.4 1.8 - -7 >32 Liver, LN 6.6 9.5 0 0
8 9 Liver, Pancreas, Bowel, LN, Other 3.3 5.3 -
-9 14 Liver 3.9 8.5 1 0
10 13 Lung, Liver, LN 7.0 22.1 0 0
11 17 Liver, Bowel, LN 2.7 5.6 2 1
12 5 LN 8.1 15.0 -
-13 7 Heart, Pancreas, Bowel, LN, Other 5.7 6.2 -
-14 9 Liver, LN 2.5 2.2 -
-15 38 Lung, Liver, LN, Bone, Other 12.4 42.3 2 0
16 12 Lung, Pancreas, LN, Bone 6.1 3.8 0 0
17 2 Bowel, LN 9.5 12.5 -
-18 >40 Liver, Bowel, LN, Bone, Other 7.5 30.7 3 1
19 9 Liver, Bowel, LN 10.6 28.7 -
-20 6 LN, Bone 13.7 28.0 -
-21 5 Liver, LN 3.5 8.4 -
-22 4 Pancreas, LN 10.5 8.3 -
-23 46 Liver, Pancreas, LN, Bone 7.8 25.0 1 0
24 17 Liver, LN 3.8 8.6 1 0
25 13 Liver, Pancreas, Other 3.5 5.0 0 0
26 20 Liver, Pancreas, LN, Bone 5.3 10.9 0 0
27 >30 Heart, Liver, LN, Bone, Other 7.7 13.9 4 0
28 >23 Liver, LN 5.8 11.2 0 0
29 4 Bowel, LN 5.2 2.9 -
-30 7 Liver 1.6 1.4 -
-31 3 LN 2.8 2.1 -
-32 5 LN 4.0 2.1 -
-33 >54 Lung, Liver, LN, Bone, Other 10.3 171 3 0
34 5 Heart, Bowel, LN 5.5 6.2 -
-35 15 Liver, LN 3.1 6.5 1 0
36 8 Liver, Bowel, LN 4.0 9.1 -
-37 15 Liver, Bone 6.2 17.2 1 1
38 >48 Liver, LN, Bone, Other 2.8 8.4 9 7
LN = lymph node, Max vs min = lesion with highest tracer uptake divided by lesion with lowest tracer uptake within a patient, MH= lesion with mild higher [18F]FDOPA uptake than other lesions, EH= lesion with extreme higher [18F]
5
mesenterium, bone and lungs were most frequently involved. Tumor burden per lesion is depicted in the Fig. S3. The median [18F]FDOPA uptake of all analyzed lesions
dif-fered per organ (p<0.05) (Fig. 3). Lesions in lung had the highest [18F]FDOPA uptake
(median 8.4) and lesions of heart the lowest (median 1.7). The median of SUV mean values measured in the background organs of all patients was 1.2 (interquartile range of 0.9-1.5) and differed per organ (p<0.05 ) (Fig. 4 and Table 3).
Effect of partial volume effect and tissue density on physiological
[
18F]FDOPA uptake
SUV peak and [18F]FDOPA uptake of all tumor lesions were correlated with their
vol-umes, r=0.20 (p= 7.33*10-8) and r=0.15 (p=6.7*10-5), respectively (Fig 5). SUV mean
of healthy lung was linearly correlated with tissue density r=0.83 (p=1.28*10-10) (Fig 6,
Table 3).
Figure 1. [18F]FDOPA uptake per tumor lesion per patient.
Tumor lesions above the inner and outer upper fences show mild or extreme high [18F]FDOPA uptake in patients
with ≥ 10 tumor lesions. Each symbol represents a single lesion. Color and kind of symbol are equal per lesion within a patient.
Figure 2. Representative [18F]FDOPA PET scan.
[18F]FDOPA uptake per lesion in patient 15. A) [18F]FDOPA PET scan B) [18F]FDOPA PET scan fused with low dose
computed tomographic scan. Written informed consent for publication of the clinical images was obtained from the patient.
Figure 3. [18F]FDOPA uptake per tumor lesion according to location in the body.
SUVpeak per tumorVOI according to location. Each symbol represents a single lesion. Kind of symbol is equal per lesion in the same kind of organ.
5
Figure 4. SUVmean per VOI according to location in the body of non-tumoral tissue.
The VOIs per organ were of the same size per organ. The bar is showing the median. Each symbol represents a single VOI. Kind of symbol is equal per VOI in the same kind of organ.
SUV=standard uptake value, VOI=volume of interest.
Table 3. Median SUV mean value of each background organ and correlation with tissue den-sity.
Organ SUV mean median
(range) Correlation SUV mean with tissue density (Pearson’s r) p-value
Kidney 2.8 (2.3 -3.3) 0.04 NS Liver 1.7 (1.3-1.9) 0.01 NS Pancreas 1.5 (1.2-1.7) -0.09 NS Myocardium 1.4 (1.2-1.6) -0.11 NS Spleen 1.2 (1.0-1.4) -0.16 NS Muscle 1.2 (1.1-1.3) -0.09 NS Bone marrow 1.1 (0.8-1.3) 0.10 NS Bloodpool 1.1 (0.9-1.2) 0.00 NS Brain 1.1 (0.8-1.3) 0.07 NS Cortical bone 0.9 (0.7-1.1) -0.17 NS Lung 0.2 (0.2-0.3) 0.83 1.28 *10-10 Total 1.2 (0.9-1.5) 0.22 5*10-6
Figure 5. Partial volume effect compared with a phantom model.
SUV peak of the lesion versus the volume of the lesion, compared with the curve of the ‘partial volume effect’ measured with a phantom.
cc= cubic centimeters, SUV=standard uptake value, VOI=volume of interest.
Figure 6. SUVmean per VOI according to tissue density.
5
Discussion
[18F]FDOPA PET scans showed interlesional tumor heterogeneity within patients with
SI-NET and identified lesions with mild or extreme high [18F]FDOPA uptake in 60% of
the evaluable patients. Interlesional tumor heterogeneity of NET tumor characteristics within patients has been shown in several pathology studies (4-6,19). Currently no stud-ies aiming to identify lesions with high tracer uptake are performed. We performed a refined analyses of tumor burden with the use of tumor to background ratio to compen-sate for differences between organs in tracer uptake, vessel density and permeability (17,20). With the knowledge that interlesional tumor heterogeneity in neuroendocrine tumors can be determined with [18F]FDOPA PET, we were interested if lesions with mild
or extreme high [18F]FDOPA uptake could be identified.
Because total tumor burden measured on [18F]FDOPA PET, correlate with 5-HIAA in
urine, a better insight in heterogeneity could potentially be used in case of systemic symptoms (10). In patients with symptomatic liver metastases, local treatment with RFA, TACE, TAE, resection or SIRT are relevant treatment modalities (9). With the selection of tumor lesions with a relatively high tumor burden, local treatments might be more effective in symptom management, than selection based on other characteristics. In general, tumor heterogeneity in neoplasms is determined by genetic variation be-tween subclones of tumor cells. As a consequence cell clones that have an advantage within a given tumor micro-environmental context survive and can expand through Darwinian selection (2). This may lead to tumor adaptation and therapeutic failure (21). Interlesional tumor heterogeneity on [18F]FDOPA PET was rudimentary analyzed in an
imaging study in 77 patients with NET. The SUV max, without correction for the SUV of the background organ, between lesions within the patient, varied up to 29-fold (10). In NET one study analyzed heterogeneity within a single tumor lesion. 141 patients with NET, of which 108 patients had a GEP-NET, underwent a Gallium-68 somatostatin ana-logue ([68Ga]SSA) PET scan to select them for radiolabeled peptide receptor therapy.
Heterogeneity within a lesion was shown to be present and entropy (a heterogeneity parameter) appeared to be predictive for progression free survival and overall survival (22). Recently, it was shown that total tumor burden in patients with SI-NET measured with [68Ga]SSA PET scan correlated with tumor markers; with 5-HIAA level in 24h
urine and with chromogranine A, however this study did not investigate heterogeneity between tumor lesions (23). In another study, in five patients the genetic heterogeneity between primary and metastatic lesions was evaluated by whole exome sequencing. This revealed a highly varying degree of genetic heterogeneity between primary lesions and hepatic metastasis (24).
In studies where sensitive molecular scans are used like in this study, many small tumor lesions could become visible. In small lesions, the partial volume effect (PVE) has to
be taken into account. Part of variability in SUV, might be explained by the PVE. This is demonstrated in the non-linear curve of our data, similar to the curve derived from the phantom model (Fig 5) (25,26). In total 90% of lesions in our study were smaller than 20 mL and therefore their SUVs could be influenced by partial volume effects (PVE). A component of PVE is tissue fraction effect (27). Healthy lung tissue has a lower den-sity because it contains air. It is found that an increase in observed SUV is associated with regions of increased lung density (27,28). In our analysis this is reflected in a linear correlation between lung density with the SUV and a lower SUV mean of lung tissue compared to SUV mean value of the other organs used for background measurement (Fig 6). This resulted in higher [18F]FDOPA uptake in lung lesions compared with lesions
in other organs (Fig 3). Several methods are developed to correct for tissue fraction effect in lung (27,28). In follow-up studies, to select lesions with substantially higher [18F]FDOPA uptake for local treatment, the tissue fraction effect and PVE have to be
taken into account.
Conclusions
In patients with SI-NET, [18F]FDOPA uptake shows considerable heterogeneity in uptake
between tumor lesions within a patient. Furthermore in some patients lesions with mild or extreme high [18F]FDOPA uptake could be identified. This suggests that [18F]FDOPA
PET scans can serve, in patients with systemic symptoms, due to an overproduction of catecholamines, to select lesions with substantially higher [18F]FDOPA uptake for local
5
References
1. Raut CP, Kulke MH, Glickman JN, Swanson RS, Ashley SW. Carcinoid tumors. Curr Probl Surg 2006;43:383-450.
2. Burrell RA, McGranahan N, Bartek J, Swanton C. The causes and consequences of genetic heterogeneity in cancer evolution. Nature 2013;501:338-45.
3. Couvelard A, Deschamps L, Ravaud P, et al. Heterogeneity of tumor prognostic markers: a reproducibility study applied to liver metastases of pancreatic endocrine tumors. Mod Pathol 2009;22:273-81.
4. Shi C, Gonzalez RS, Zhao Z, et al. Liver metastases of small intestine neuroendocrine tumors: Ki-67 heterogeneity and World Health Organization grade discordance with primary tumors. Am J Clin Pathol 2015;143:398-404.
5. Richards-Taylor S, Tilley C, Jaynes E, et al. Clinically significant differences in Ki-67 pro-liferation index between primary and metastases in resected pancreatic neuroendocrine tumors. Pancreas 2017;46:1354-8.
6. Charoenpitakchai M, Liu E, Zhao Z, et al. In liver metastases from small intestinal neuroen-docrine tumors, SSTR2A expression is heterogeneous. Virchows Arch 2017;470:545-52. 7. Pavel M, Baudin E, Couvelard A, et al. ENETS Consensus Guidelines for the
manage-ment of patients with liver and other distant metastases from neuroendocrine neoplasms of foregut, midgut, hindgut, and unknown primary. Neuroendocrinology 2012;95:157-76. 8. Kennedy A, Bester L, Salem R, et al. Role of hepatic intra-arterial therapies in metastatic
neuroendocrine tumours (NET): guidelines from the NET-Liver-Metastases Consensus Conference. HPB (Oxford) 2015;17:29-37.
9. van Hazel GA, Heinemann V, Sharma NK, et al. SIRFLOX: Randomized phase III trial comparing first-line mFOLFOX6 (plus or minus bevacizumab) versus mFOLFOX6 (plus or minus bevacizumab) plus selective internal radiation therapy in patients with metastatic colorectal cancer. J Clin Oncol 2016;34:1723-31.
10. Fiebrich HB, de Jong JR, Kema IP, et al. Total 18F-dopa PET tumour uptake reflects
meta-bolic endocrine tumour activity in patients with a carcinoid tumour. Eur J Nucl Med Mol Imaging 2011;38:1854-61.
11. Kema IP, de Vries EG, Schellings AM, Postmus PE, Muskiet FA. Improved diagnosis of carcinoid tumors by measurement of platelet serotonin. Clin Chem 1992;38:534-40. 12. Luurtsema G, Boersma HH, Schepers M, et al. Improved GMP-compliant multi-dose
pro-duction and quality control of 6-[18F]fluoro-L-DOPA. EJNMMI Radiopharm Chem 2017;1:7.
13. Koopmans KP, de Vries EG, Kema IP, et al. Staging of carcinoid tumours with 18F-DOPA
PET: a prospective, diagnostic accuracy study. Lancet Oncol 2006;7:728-34.
14. Hudson HM, Larkin RS. Accelerated image reconstruction using ordered subsets of projec-tion data. IEEE Trans Med Imaging 1994;13:601-9.
15. van Asselt SJ, Oosting SF, Brouwers AH, et al. Everolimus reduces 89Zr-bevacizumab
tumor uptake in patients with neuroendocrine tumors. J Nucl Med 2014;55:1087-92. 16. Koopmans KP, Neels OC, Kema IP, et al. Improved staging of patients with carcinoid and
islet cell tumors with 18F-dihydroxy-phenyl-alanine and 11C-5-hydroxy-tryptophan positron
emission tomography. J Clin Oncol 2008;26:1489-95.
17. Frings V, van Velden FH, Velasquez LM, et al. Repeatability of metabolically active tumor volume measurements with FDG PET/CT in advanced gastrointestinal malignancies: a multicenter study. Radiology 2014;273:539-48.
18. National Electrical Manufacturers Association. NEMA Standards publication NU 2-2007, Performance measurements of positron emission tomographs. 2007.
19. Francis JM, Kiezun A, Ramos AH, et al. Somatic mutation of CDKN1B in small intestine neuroendocrine tumors. Nat Genet 2013;45:1483-6.
20. Bol K, Haeck JC, Groen HC, et al. Can DCE-MRI explain the heterogeneity in radiopep-tide uptake imaged by SPECT in a pancreatic neuroendocrine tumor model? PLoS One 2013;8:e77076.
21. Gerlinger M, Rowan AJ, Horswell S, et al. Intratumor heterogeneity and branched evolution revealed by multiregion sequencing. N Engl J Med 2012;366:883-92.
22. Werner RA, Lapa C, Ilhan H, et al. Survival prediction in patients undergoing radionu-clide therapy based on intratumoral somatostatin-receptor heterogeneity. Oncotarget 2017;8:7039-49.
23. Tirosh A, Papadakis GZ, Millo C, et al. Association between neuroendocrine tumors bio-markers and primary tumor site and disease type based on total (68)Ga-DOTATATE-Avid tumor volume measurements. Eur J Endocrinol 2017;176:575-82.
24. Walter D, Harter PN, Battke F, et al. Genetic heterogeneity of primary lesion and metastasis in small intestine neuroendocrine tumors. Sci Rep 2018;8:3811.
25. Hoetjes NJ, van Velden FH, Hoekstra OS, et al. Partial volume correction strategies for quantitative FDG PET in oncology. Eur J Nucl Med Mol Imaging 2010;37:1679-87. 26. Hoffman EJ, Huang SC, Phelps ME. Quantitation in positron emission computed
tomogra-phy: 1. Effect of object size. J Comput Assist Tomogr 1979;3:299-308.
27. Holman BF, Cuplov V, Millner L, et al. Improved correction for the tissue fraction effect in lung PET/CT imaging. Phys Med Biol 2015;60:7387-402.
28. Lambrou T, Groves AM, Erlandsson K, et al. The importance of correction for tissue fraction effects in lung PET: preliminary findings. Eur J Nucl Med Mol Imaging 2011;38:2238-46.
5
All 18F-DOPA PET/CT scans
made in adults in the UMCG from 4-2-2014 to 1-4-2015. (Excluding PET only camera)
n=182 scans (n=160 patients)
Only patients with SI-NET
n=61 scans (n=48 patients)
Only patients wtih ≥ 1 positive
lesion on 18F-DOPA PET/scan
n=93 scans (n=77 patients)
Exclusion ≤ 1 positive lesion on 18F-DOPA PET/scan:
n=89 scans (n= 83 patients)
Only patients with CT and 18
F-DOPA PET scan
n=48 scans -> (n=38 patients)
Exclusion patients with primary tumor of pancreas or bronchopulmonal
n= 32 scans (n=29 patients)
Exclusion no CT-scan was made within 6 months
before or after the 18F-DOPA PET/scan
n=13 scans (n=10 patients)
Organ N um be r of le si on s Bone Bowe l Heart Liver Lung LN/m esen terium Panc reas Other 0 100 200 300 400 500 PET lesions CT inconclusive or malignant lesions Organ Nr o f l es io ns Bone Bowe l Heart Liver Lung LN / m esen terium Panc reas Other 0 100 200 300 400 500 PET lesions Malignant CT lesions
5
Patients 18 F-DO PA * vo lu m e (m L) 1 2 3 4 5 6 7 8 9 1011121314151617181920212223242526272829303132333435363738 0 1000 2000 3000 4000 5000Figure S3. Tracer uptake multiplied by volume per lesion per patient.
